src/labdaps/config.py

from pathlib import Path

# Paths
BASE_DIR = Path(__file__).parent.parent.parent
DOCS_DIR = BASE_DIR / "docs"
CHROMA_DIR = BASE_DIR / "data" / "chroma_db"

# ChromaDB
COLLECTION_NAME = "labdaps_sinal_alerta"

# Embedding model (local, multilingual, Portuguese-friendly)
EMBEDDING_MODEL = "intfloat/multilingual-e5-small"
EMBED_QUERY_PREFIX = "query: "
EMBED_PASSAGE_PREFIX = "passage: "

# Chunking
CHUNK_SIZE_CHARS = 2400
CHUNK_OVERLAP_CHARS = 450
MIN_CHUNK_CHARS = 200

# Retrieval
TOP_K_CANDIDATES = 10
TOP_K_FINAL = 5

# Groq / Llama
GROQ_MODEL = "llama-3.3-70b-versatile"
MAX_HISTORY_MESSAGES = 6